Data Visualisation¶

Table of Contents¶

Importing Libraries¶

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import warnings
import folium
import os
import branca.colormap
from collections import defaultdict
import geopandas as gpd
from shapely.geometry import Point, Polygon
from folium.plugins import MarkerCluster
from folium.plugins import HeatMap
import branca.colormap as cm
import gmaps
from statsmodels.formula.api import ols
warnings.filterwarnings("ignore")

Importing Data¶

filenames = ['df_2016to2019sales.csv', 'df_yearly_summary.csv', 'df_crime.csv', 'mean_crime.csv',
             'df_schooldata.csv', 'df_rat.csv', 'df_ratperland.csv', 'schools_land.csv', 'population.csv']

visualisation_dfs = []
for i in filenames:
    visualisation_dfs.append(pd.read_csv('C:/Users/ChengYao/Desktop/DAO2702/Project/Clean Data/' + i))

Property Price Trend From 2007 to 2019¶

Sale Price vs Years (per borough) for One/Two/Three Family Homes¶

def plot_lines_years(data, boroughs, years, title=''):
    sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})
    years_str = list(map(str, years)) #change all years into a string and then add into a list
    plt.figure(figsize=(15, 10)) #set figure size
    for i in boroughs: 
        sns.lineplot(y='AVERAGE SALE PRICE',
                     x='YEAR',
                     data=data.groupby(['BOROUGH', 
                                        'YEAR']).mean().loc[i].reset_index(), #this group by will filter by the borough and get the average price per year
                     label=i, #label it as the borough name
                     lw=3, marker='o', ms=10)
    plt.title("Average Sale Price from 2007 - 2019 {}".format(title),
             pad=10, fontsize=18) #title
    plt.xticks(years, years_str) #plot the x-axis years but with labels as str format
    plt.show() #plot graph out

buildingcategories = visualisation_dfs[1]['BUILDING CLASS CATEGORY'].unique()
yearlysummarydata = visualisation_dfs[1]

onehomedata = yearlysummarydata.loc[(yearlysummarydata['BUILDING CLASS CATEGORY'] == buildingcategories[0])]
twohomedata = yearlysummarydata.loc[(yearlysummarydata['BUILDING CLASS CATEGORY'] == buildingcategories[1])]
threehomedata = yearlysummarydata.loc[(yearlysummarydata['BUILDING CLASS CATEGORY'] == buildingcategories[2])]

boroughs = yearlysummarydata['BOROUGH'].unique()
years = yearlysummarydata['YEAR'].unique()

One/Two/Three Family Homes Graphs¶

plot_lines_years(onehomedata,boroughs,years, title = 'of One Family Homes' )

plot_lines_years(twohomedata,boroughs,years,title = "of Two Family Homes")

plot_lines_years(threehomedata,boroughs,years, title = "of Three Family Homes")

plot_lines_years(yearlysummarydata, boroughs,years,title = 'of each Borough')

Percentage Increase in Average Sale Price per Borough from 2007 to 2019¶

def plotdouble(borough,barcolor,leftcolor,linecolor):
    data = visualisation_dfs[1]
    data = data.loc[(data['BOROUGH'] == borough)]
    data1 = data.groupby(['YEAR']).mean().reset_index()
    data1['% INCREASE'] = data1['% INCREASE'] * 100

    norm = plt.Normalize(data1['AVERAGE SALE PRICE'].min(), data1['AVERAGE SALE PRICE'].max())
    sm = plt.cm.ScalarMappable(cmap=barcolor, norm=norm)
    sm.set_array([])

    sns.set(style='white', font_scale=1.3, font="calibri", 
            rc={'axes.facecolor':'#d8d8d8'})
    fig,ax1 = plt.subplots(figsize = (15,10))
    sns.barplot(x = 'YEAR', y = 'AVERAGE SALE PRICE', data = data1 , hue = 'AVERAGE SALE PRICE', dodge = False ,palette = barcolor, ax = ax1, ci = None)
    ax2 = ax1.twinx()

    sns.lineplot(x = data1['YEAR'].index, y= data1['% INCREASE'], ax = ax2, color = linecolor)

    ax1.get_legend().remove()
    ax1.figure.colorbar(sm, pad = 0.1,fraction = 0.046)

    ax1.set_xlabel('Year',fontsize = 18)
    ax1.set_ylabel('Average Sale Price',fontsize = 18, color = leftcolor)
    ax2.set_ylabel('Percentage Change',fontsize = 18,color = linecolor)
    plt.title('Average Sale Price & Percentage Increase of ' + borough + ' from 2007 to 2019',fontsize = 20)
    plt.show()

plotdouble('Bronx','Reds', 'Red', 'Black')

plotdouble('Brooklyn','Blues','Blue','Red')

plotdouble('Queens','Greens','Green','Red')

plotdouble('Staten Island','Purples','Purple','Red')

plotdouble('Manhattan','Greys','Grey','Red')

Sale Price vs Borough Barplot¶

df_4yearscombined = visualisation_dfs[0]

#def sale_borough(data):
sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})
years = [2016, 2017, 2018, 2019]
data_year_wise = []
for i in years:
    data_year_wise.append(df_4yearscombined.loc[df_4yearscombined['YEAR'] == i, ['SALE PRICE', 'BOROUGH']])
fig, ax = plt.subplots(ncols=2, nrows=2, figsize=(15, 10))
for i, j in enumerate(ax.reshape(-1)):
    sns.barplot(x='BOROUGH',
                y='SALE PRICE',
                data=data_year_wise[i],
                ax=j,
                estimator=np.mean,
                ci=None)
    j.title.set_text(str(years[i]))
    j.set_ylabel('Mean Sale Price')
    j.set_xlabel('Borough')
fig.tight_layout(pad=2.0)
plt.suptitle("Mean Sale Price vs Borough", fontsize=20, y=1.01)
plt.show()

Sale Price vs Borough Boxplot¶

sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})
years = [2016, 2017, 2018, 2019]
data_year_wise = []
for i in years:
    data_year_wise.append(df_4yearscombined.loc[df_4yearscombined['YEAR'] == i, ['SALE PRICE', 'BOROUGH']])
fig, ax = plt.subplots(ncols=2, nrows=2, figsize=(15, 10))
for i, j in enumerate(ax.reshape(-1)):
    sns.boxplot(x='BOROUGH',
                y='SALE PRICE',
                data=data_year_wise[i],
                ax=j)
    j.title.set_text(str(years[i]))
    j.set_ylabel('Sale Price')
    j.set_xlabel('Borough')
fig.tight_layout(pad=2.0)
plt.suptitle("Sale Price vs Borough", fontsize=20, y=1.01)
plt.show()

bronx = df_4yearscombined[df_4yearscombined['BOROUGH'] == 'Bronx']
brooklyn = df_4yearscombined[df_4yearscombined['BOROUGH'] == 'Brooklyn']
queens = df_4yearscombined[df_4yearscombined['BOROUGH'] == 'Queens']
statenisland = df_4yearscombined[df_4yearscombined['BOROUGH'] == 'Staten Island']
manhattan = df_4yearscombined[df_4yearscombined['BOROUGH'] == 'Manhattan']

bronx = bronx['SALE PRICE'].describe()
brooklyn = brooklyn['SALE PRICE'].describe()
queens = queens['SALE PRICE'].describe()
statenisland = statenisland['SALE PRICE'].describe()
manhattan = manhattan['SALE PRICE'].describe()

df10 = pd.DataFrame()
df10 = pd.concat([df10,bronx,brooklyn,queens,statenisland,manhattan],axis = 1,ignore_index = True)
df10.columns= ['Bronx','Brooklyn','Queens','Staten Island','Manhattan']

df10

sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})

data1 = df_4yearscombined.loc[(df_4yearscombined["SALE PRICE"] < 9600000)]
#data1 = data1[data1['BOROUGH'].isin(['Bronx','Brooklyn','Queens','Staten Island'])]
years = [2016, 2017, 2018, 2019]
data_year_wise = []
for i in years:
    data_year_wise.append(data1.loc[data1['YEAR'] == i, ['SALE PRICE', 'BOROUGH']])
fig, ax = plt.subplots(ncols=2, nrows=2, figsize=(15, 10))
for i, j in enumerate(ax.reshape(-1)):
    sns.boxplot(x='BOROUGH',
                y='SALE PRICE',
                data=data_year_wise[i],
                ax=j)
    j.title.set_text(str(years[i]))
    j.set_ylabel('Sale Price')
    j.set_xlabel('Borough')
fig.tight_layout(pad=2.0)
plt.suptitle("Sale Price vs Borough", fontsize=20, y=1.01)
plt.show()

sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})

#data2 = df_4yearscombined.loc[(df_4yearscombined["SALE PRICE"] < 20000000)]
data2 = df_4yearscombined[df_4yearscombined['BOROUGH'].isin(['Manhattan'])]
years = [2016, 2017, 2018, 2019]
data_year_wise = []
for i in years:
    data_year_wise.append(data2.loc[data2['YEAR'] == i, ['SALE PRICE', 'BOROUGH']])
fig, ax = plt.subplots(ncols=2, nrows=2, figsize=(15, 10))
for i, j in enumerate(ax.reshape(-1)):
    sns.boxplot(x='BOROUGH',
                y='SALE PRICE',
                data=data_year_wise[i],
                ax=j)
    j.title.set_text(str(years[i]))
    j.set_ylabel('Sale Price')
    j.set_xlabel('Borough')
fig.tight_layout(pad=2.0)
plt.suptitle("Sale Price vs Borough", fontsize=20, y=1.01)
plt.show()

Sale Price vs Land Square Feet Scatterplot¶

#def sale_land(data):
sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})
years = [2016, 2017, 2018, 2019]
colors_ = ['#cd025c', '#f5ab60', '#743fc5', '#243e1f']
data_year_wise = []
for i in years:
    data_year_wise.append(df_4yearscombined.loc[df_4yearscombined['YEAR'] == i, ['SALE PRICE', 'LAND SQUARE FEET']])
fig, ax = plt.subplots(ncols=2, nrows=2, figsize=(15, 10))
for i, j in enumerate(ax.reshape(-1)):
    sns.regplot(x='LAND SQUARE FEET',
                y='SALE PRICE',
                data=data_year_wise[i],
                ax=j,
                color=colors_[i])
    j.title.set_text(str(years[i]))
    j.set_ylabel('Sale Price')
    j.set_xlabel('LAND SQUARE FEET')
fig.tight_layout(pad=2.0)
plt.suptitle("Sale Price vs Land Square Feet", fontsize=20, y=1.01)
plt.show()

Sale Price vs Tax Class¶

sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})
years = [2016, 2017, 2018, 2019]
data_year_wise = []
for i in years:
    data_year_wise.append(df_4yearscombined.loc[df_4yearscombined['YEAR'] == i, ['SALE PRICE', 'TAX CLASS']])
fig, ax = plt.subplots(ncols=2, nrows=2, figsize=(15, 10))
for i, j in enumerate(ax.reshape(-1)):
    sns.barplot(x='TAX CLASS',
                y='SALE PRICE',
                data=data_year_wise[i],
                ax=j,
                order = ['1','1B','2A','2B','2','4'],
                estimator=np.mean,
                ci=None)
    j.title.set_text(str(years[i]))
    j.set_ylabel('Mean Sale Price')
    j.set_xlabel('Tax Class')
fig.tight_layout(pad=2.0)
plt.suptitle("Mean Sale Price vs Tax Class", fontsize=20, y=1.01)
plt.show()

Sale Price vs Year Built¶

sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})
years = [2016, 2017, 2018, 2019]
data_year_wise = []
for i in years:
    data_year_wise.append(df_4yearscombined.loc[df_4yearscombined['YEAR'] == i, ['SALE PRICE', 'YEAR BUILT (Bins)']])
fig, ax = plt.subplots(ncols=2, nrows=2, figsize=(15, 10))
for i, j in enumerate(ax.reshape(-1)):
    sns.barplot(x='YEAR BUILT (Bins)',
                y='SALE PRICE',
                data=data_year_wise[i],
                order = ['< 1900', '1900 - 1924', '1925 - 1949', '1950 - 1974','1975 - 1999','2000 - 2019'],
                ax=j,
                estimator = np.mean,
                ci=None)
    j.title.set_text(str(years[i]))
    j.set_ylabel('Mean Sale Price', labelpad=10)
    j.set_xlabel('Year Built (Bins)', labelpad=10)
    j.set_xticklabels(j.get_xticklabels(), rotation=15)
fig.tight_layout(pad=2.0)
plt.suptitle("Mean Sale Price vs Year Built (Bins)", fontsize=20, y=1.01)
plt.show()

¶

Sale Price vs Neighbourhoods in Each Borough¶

neighbourhood = df_4yearscombined[['SALE PRICE', 'YEAR','NEIGHBORHOOD','BOROUGH']]
neighbourhood = neighbourhood.loc[(neighbourhood['YEAR'] == 2019)]
neighbourhood = neighbourhood.sort_values(['BOROUGH'], ascending = True)

boroughs = neighbourhood['BOROUGH'].unique() #unique values given are in order of the datafile and we have checked thus reason why we sorted earlier

data_borough_wise = []
for i in boroughs:
    data_borough_wise.append(neighbourhood.loc[neighbourhood['BOROUGH'] == i, ['SALE PRICE', 'NEIGHBORHOOD']])


def neighbourhood_plot(data,i,title): #function to plot the neighbourhood barplots
    sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
            rc={'axes.facecolor':'#d8d8d8'})

    fig, ax = plt.subplots(ncols=1, nrows=1, figsize=(15, 10))

    sns.barplot(x='NEIGHBORHOOD',
                y='SALE PRICE',
                data=data[i],
                estimator = np.mean,
                ci=None)

    ax.set_ylabel('Mean Sale Price', labelpad=10)
    ax.set_xlabel('Neighbourhoods', labelpad=10)
    ax.set_xticklabels(ax.get_xticklabels(), rotation=90, fontsize = 14)
    fig.tight_layout(pad=2.0)
    plt.suptitle("Mean Sale Price vs Neighbourhoods in " + title , fontsize=20, y=1.01)
    meanvalue = data[i]['SALE PRICE'].mean()
    plt.axhline(y=meanvalue)
    ax.legend(['Mean Sale Price in ' + title])
    plt.show()

neighbourhood_plot(data_borough_wise,0,"Bronx")

neighbourhood_plot(data_borough_wise,1,"Brooklyn")

neighbourhood_plot(data_borough_wise,2,"Manhattan")

neighbourhood_plot(data_borough_wise,3,"Queens")

neighbourhood_plot(data_borough_wise,4,"Staten Island")

¶

Public High School Data¶

¶

highschooldata = visualisation_dfs[7]
highschooldata['Schools per Land Size'] = highschooldata['Number of Schools']/highschooldata['Land Size']

highschooldata

Number of Public High Schools per Borough Barplot¶

plt.figure(figsize=(15, 8))
sns.barplot(x='Borough',
            y='Number of Schools',
            data=highschooldata)
plt.title("Barplot showing number of public high schools for each Borough",
         pad=10,
         fontsize=20)
plt.show()

¶

Schools per km^2 per Borough Barplot¶

plt.figure(figsize=(15, 8))
sns.barplot(x='Borough',
            y='Schools per Land Size',
            data=highschooldata)
plt.title("Barplot showing Schools per km^2 for each Borough",
         pad=10,
         fontsize=20)
plt.show()

pop = visualisation_dfs[8]
pop1 = pop.loc[(pop['Age Group'] == 'School-Age (5 to17) ')]
pop1 = pop1[['Borough','2020']]
pop1.columns = ['Borough','Student Population']

highschooldata = pd.merge(highschooldata,pop1,on ='Borough')

highschooldata['Schools per Student Population'] = highschooldata['Number of Schools']/highschooldata['Student Population']

highschooldata

¶

Schools per Student Population per Borough Barplot¶

plt.figure(figsize=(15, 8))
sns.barplot(x='Borough',
            y='Schools per Student Population',
            data=highschooldata)
plt.title("Barplot showing Schools per Student Population for each Borough",
         pad=10,
         fontsize=20)
plt.show()

¶

Average SAT Scores per Borough Boxplot¶

satscores = visualisation_dfs[4]


plt.figure(figsize=(15, 10))
sns.boxplot(x='Borough',
        y='Average SAT Score',
        data=satscores, palette='Paired', order = ['Bronx','Brooklyn','Manhattan','Queens','Staten Island'])
plt.title("Average SAT scores in each Borough", pad=10, fontsize=20)
plt.show()

lat_long = visualisation_dfs[4]
lat_long = lat_long[['Latitude',  'Longitude', 'School Name', 'Borough']]
lat_long = lat_long.dropna() #incase there is NA

borough_poly = gpd.read_file("C:/Users/ChengYao/Desktop/DAO2702/Project/Clean Data/Borough Boundaries.geojson")
lat_long.head(3)

¶

Clustered Marker Map of Schools in NYC¶

map_markers = folium.Map(location=[40.75321, -73.99786], zoom_start=10)
marker_cluster_school = MarkerCluster().add_to(map_markers)

for i in range(0, len(lat_long)):
    folium.Marker(lat_long.iloc[i, :2].tolist(), 
                  popup=lat_long.iloc[i,2] + '<br>' + '<b>' + lat_long.iloc[i,3] + '<b>')\
                .add_to(marker_cluster_school)
folium.GeoJson(borough_poly['geometry'][0]).add_to(map_markers)
folium.GeoJson(borough_poly['geometry'][1]).add_to(map_markers)
folium.GeoJson(borough_poly['geometry'][2]).add_to(map_markers)
folium.GeoJson(borough_poly['geometry'][3]).add_to(map_markers)
folium.GeoJson(borough_poly['geometry'][4]).add_to(map_markers)

map_markers

¶

Rat Sightings¶

rats = visualisation_dfs[5]


rats1 = rats['Borough'].value_counts().reset_index() #calculate number of sightings per borough
rats1.columns = ['Borough', 'Total Number of Rat Sightings']

Number of Rat Sightings per Borough Barplot¶

plt.figure(figsize=(15, 8))
sns.barplot(x='Borough',
            y='Total Number of Rat Sightings',
            data=rats1, palette='Set2')
plt.title("Barplot showing Total Number of Rat Sightings for each Borough", 
          pad=10, fontsize=20)
plt.show()

¶

Average Number of Sightings per km^2 per Borough Barplot¶

rats2 = visualisation_dfs[6]

plt.figure(figsize=(15, 8))
sns.barplot(x='Borough',
            y='Average No. of Sightings/km^2',
            order = ['Brooklyn','Manhattan','Bronx','Queens','Staten Island'],
            data=rats2, palette='husl')
plt.title("Barplot showing Average No. of Sightings per km^2 for each Borough", 
          pad=10, fontsize=20)
plt.show()

Rat Sightings Heatmap in NYC¶

m = folium.Map([40.75321, -73.99786], zoom_start=10)

heat_arr = rats.dropna(subset=['Latitude', 'Longitude'])
heat_arr = heat_arr[['Latitude', 'Longitude']]

steps=20
colormap = branca.colormap.linear.YlOrRd_09.scale(0, 1).to_step(steps)
gradient_map=defaultdict(dict)
for i in range(steps):
    gradient_map[1/steps*i] = colormap.rgb_hex_str(1/steps*i)
colormap.add_to(m) #add color bar at the top of the map

# plot heatmap
m.add_children(HeatMap(heat_arr, radius=9,gradient = gradient_map))
folium.GeoJson(borough_poly['geometry'][0]).add_to(m)
folium.GeoJson(borough_poly['geometry'][1]).add_to(m)
folium.GeoJson(borough_poly['geometry'][2]).add_to(m)
folium.GeoJson(borough_poly['geometry'][3]).add_to(m)
folium.GeoJson(borough_poly['geometry'][4]).add_to(m)
m

Rat Sightings Google HeatMap¶

gmaps.configure(api_key = 'AIzaSyCPDB6lIcli7R-haaCtIdnVIlFhs6fpW7M')  #my API key
m1 = gmaps.figure(center = (40.730610, -73.935242), zoom_level = 10)
m1.add_layer(gmaps.heatmap_layer(heat_arr))
m1

###The graph doesn't show in HTML but it will run in Jupyter¶

¶

Population Data¶

Next we want to see the population distribution in each borough and the projected population over the next 2 decades¶

pop = visualisation_dfs[8]
pop1 = pop.iloc[:,[0,1,2,4,6,8]]

def plotpopulation(lst,j,title = ''):
    barWidth = 0.25
    
    bars1 = pop1.iloc[lst[0],2:].tolist()
    bars2 = pop1.iloc[lst[1],2:].tolist()
    bars3 = pop1.iloc[lst[2],2:].tolist()

    r1 = np.arange(len(bars1))
    r2 = [x + barWidth for x in r1]
    r3 = [x + barWidth for x in r2]

    j.bar(r1, bars1, color='#6497b1', width=barWidth, edgecolor='white', label='Total Population')
    j.bar(r2, bars2, color='#35b899', width=barWidth, edgecolor='white', label='Student Population (5 - 17)')
    j.bar(r3, bars3, color='#b3c0e2', width=barWidth, edgecolor='white', label='Elderly ( >= 65 )')
    
    j.title.set_text(title)
    j.set_xlabel('Years')
    j.set_ylabel('Number in Population')
    j.set_xticklabels(['dummy','       2010','       2020','       2030','       2040'])

Population Growth Per Borough Grouped Barplot¶

fig, ax = plt.subplots(ncols=3, nrows=2, figsize=(15, 10))
fig.delaxes(ax[1,2])
ax = ax.reshape(-1)

bronx,brooklyn,manhattan,queens,statenisland = [1,7,13],[2,8,14],[3,9,15],[4,10,16],[5,11,17]
total = [bronx,brooklyn,manhattan,queens,statenisland]
boroughslst = boroughs.tolist() #defined earlier above

for i,j in enumerate(ax[:-1]):
    plotpopulation(total[i],j,title = boroughslst[i])
    

plt.suptitle("Grouped Barplot showing Population Growth from 2010 to 2040", fontsize=20)
plt.tight_layout()
plt.subplots_adjust(top = 0.9)
ax[4].legend(loc='center left', bbox_to_anchor=(1.2, 0.5),prop={'size':20})
plt.show()

¶

Crime Data¶

crimes = visualisation_dfs[2]
meancrimes = visualisation_dfs[3]

Total Number of Crimes from 1990 to 2018 for all 5 Boroughs Lineplot¶

sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})
years = crimes['Year'].unique()
years_str = list(map(str, years))
plt.figure(figsize=(15,10))
for i in boroughs:
    sns.lineplot(y = 'Index Total', x = 'Year',
                 data = crimes.groupby(['Borough','Year']).mean().loc[i].reset_index(), 
                 label=i,lw=3,marker='o',ms=10) ##this doesn't get the mean since every year only has one value. I did this just to get a value out
plt.title('Total Number of Crimes from 1990 to 2018 in all 5 Boroughs ',pad=10, fontsize=18) #title
plt.xticks(years, years_str,rotation = 45) #plot the x-axis years but with labels as str format
plt.ylabel('Total Number of Crimes')
plt.show() #plot graph out

Number of crimes per population in all 5 boroughs¶

sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})
years = crimes['Year'].unique()
years_str = list(map(str, years))
plt.figure(figsize=(15,10))
for i in boroughs:
    sns.lineplot(y = 'Mean No. of Crime', x = 'Year',
                 data = meancrimes.groupby(['Borough','Year']).mean().loc[i].reset_index(), 
                 label=i,lw=3,marker='o',ms=10) ##this doesn't get the mean since every year only has one value. I did this just to get a value out
plt.title('Number of Crimes per Population from 1990 to 2018 in all 5 Boroughs ',pad=10, fontsize=18) #title
plt.xticks(years, years_str,rotation = 45) #plot the x-axis years but with labels as str format
plt.ylabel('Number of Crimes per Population')
plt.show() #plot graph out

landsize = highschooldata[['Borough','Land Size']]

meancrimes2 = pd.merge(crimes,landsize,on ='Borough')

meancrimes3 = meancrimes2[['Borough','Index Total','Land Size','Year']]

meancrimes3['No. of Crimes per km^2'] = meancrimes3['Index Total']/meancrimes3['Land Size']

Number of Crimes per area in all 5 boroughs¶

sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})
years = crimes['Year'].unique()
years_str = list(map(str, years))
plt.figure(figsize=(15,10))
for i in boroughs:
    sns.lineplot(y = 'No. of Crimes per km^2', x = 'Year',
                 data = meancrimes3.groupby(['Borough','Year']).mean().loc[i].reset_index(), 
                 label=i,lw=3,marker='o',ms=10) ##this doesn't get the mean since every year only has one value. I did this just to get a value out
plt.title('Number of Crimes per km^2 from 1990 to 2018 in all 5 Boroughs ',pad=10, fontsize=18) #title
plt.xticks(years, years_str,rotation = 45) #plot the x-axis years but with labels as str format
plt.ylabel('Number of Crimes per km^2')
plt.show() #plot graph out

crimes2018 = crimes.loc[(crimes['Year'] == 2018)]

crimes2018.reset_index(drop = True)
boroughnamecrimes = crimes2018['Borough'].unique()

crimes2018list = []
typeofcrimes = crimes2018.columns.tolist()
typeofcrimes = typeofcrimes[6:14]

for i in range(0,5):
    df = crimes2018.iloc[i,6:14].tolist()
    indextotal = [crimes2018.iloc[i,4].tolist()] * 8
    df = pd.DataFrame(list(zip(df,typeofcrimes,indextotal)),columns =['Number of Crimes','Type of Crime','Index Total'])
    df['Proportion of Type of Crime'] = df['Number of Crimes']/df['Index Total']
    df = df.drop(4,axis = 0)
    df = df.reset_index(drop = True)
    crimes2018list.append(df)

crimes2018list[0]

def crimetype(data,boroughdata):
    sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
            rc={'axes.facecolor':'#d8d8d8'})
    fig, ax = plt.subplots(ncols=2, nrows=1, figsize=(15,10))

    for i,j in enumerate(ax.reshape(-1)):
        sns.barplot(x='Type of Crime',
                y='Proportion of Type of Crime',
                data=data[i], palette='Set2',ax = j)
        j.set_xticklabels(j.get_xticklabels(), rotation=70)
        j.set_title(boroughdata[i],fontsize = 20)
        j.set_ylabel('Proprtion of Crimes',fontsize = 20)
        j.set_xlabel('Type of Crimes',fontsize = 20)
    fig.tight_layout(pad=2.0)
    plt.suptitle("Proprtion of Different Types of Crimes In 2018", fontsize=20, y=1.01)
    plt.show()

Proportion of Types of Crimes in 2018¶

crimetype(crimes2018list[0:2],boroughnamecrimes[0:2])

crimetype(crimes2018list[2:4],boroughnamecrimes[2:4])

sns.set(style='darkgrid', font_scale=1.3, font="calibri", 
        rc={'axes.facecolor':'#d8d8d8'})
plt.figure(figsize=(8,9))


sns.barplot(x='Type of Crime',y='Proportion of Type of Crime',data=crimes2018list[4], palette='Set2')
plt.title('Staten Island',fontsize = 18)
plt.suptitle("Proprtion of Different Types of Crimes In 2018", fontsize=20, y=1.01)
plt.xticks(rotation = 70)
plt.xlabel('Type of Crimes',fontsize = 20)
plt.ylabel('Proprtion of Crimes', fontsize = 20)
plt.show()

	Bronx	Brooklyn	Queens	Staten Island	Manhattan
25%	3.950000e+05	5.900000e+05	4.650000e+05	3.950000e+05	1.300000e+06
50%	5.250000e+05	9.000000e+05	6.680000e+05	5.168000e+05	3.707500e+06
75%	7.000000e+05	1.450000e+06	9.062740e+05	6.516800e+05	9.600000e+06
count	1.451900e+04	3.464300e+04	4.490400e+04	2.091700e+04	6.068000e+03
max	1.334000e+08	8.696129e+08	4.750000e+08	5.200000e+07	2.397502e+09
mean	1.128478e+06	1.676580e+06	1.021438e+06	5.814295e+05	1.605500e+07
min	1.006000e+03	1.500000e+03	1.041000e+03	1.162000e+03	1.800000e+03
std	3.873449e+06	7.345476e+06	4.816893e+06	8.371688e+05	7.725789e+07

	Borough	Number of Schools	Land Size	Schools per Land Size
0	Bronx	118	109.04	1.082172
1	Brooklyn	121	183.42	0.659688
2	Manhattan	106	59.13	1.792660
3	Queens	80	281.09	0.284606
4	Staten Island	10	151.18	0.066146

	Borough	Number of Schools	Land Size	Schools per Land Size	Student Population	Schools per Student Population
0	Bronx	118	109.04	1.082172	259013	0.000456
1	Brooklyn	121	183.42	0.659688	441049	0.000274
2	Manhattan	106	59.13	1.792660	162931	0.000651
3	Queens	80	281.09	0.284606	341062	0.000235
4	Staten Island	10	151.18	0.066146	78759	0.000127

	Number of Crimes	Type of Crime	Index Total	Proportion of Type of Crime
0	91.0	Murder	35355.0	0.002574
1	662.0	Rape	35355.0	0.018724
2	3213.0	Robbery	35355.0	0.090878
3	8548.0	Aggravated Assault	35355.0	0.241776
4	2095.0	Burglary	35355.0	0.059256
5	19434.0	Larceny	35355.0	0.549682
6	1312.0	Motor Vehicle Theft	35355.0	0.037109

	Latitude	Longitude	School Name	Borough
0	40.75321	-73.99786	Clinton School Writers and Artists	Manhattan
1	40.86605	-73.92486	Inwood Early College for Health and Informatio...	Manhattan
2	40.71873	-73.97943	New Explorations into Science, Technology and ...	Manhattan

Data Visualisation¶

Table of Contents¶

Property Prices Data From 2007 to 2019 Visualisation¶

School Data Visualisation¶

Rat Sightings Data Visualisation¶

Population Data Visualisation¶

Crime Data From 1990 to 2018 Visualisation¶

Importing Libraries¶

Importing Data¶

Property Price Trend From 2007 to 2019¶

Sale Price vs Years (per borough) for One/Two/Three Family Homes¶

One/Two/Three Family Homes Graphs¶

Percentage Increase in Average Sale Price per Borough from 2007 to 2019¶

Sale Price vs Borough Barplot¶

Sale Price vs Borough Boxplot¶

Sale Price vs Land Square Feet Scatterplot¶

Sale Price vs Tax Class¶

Sale Price vs Year Built¶

Sale Price vs Neighbourhoods in Each Borough¶

Public High School Data¶

Number of Public High Schools per Borough Barplot¶

Schools per km^2 per Borough Barplot¶

Schools per Student Population per Borough Barplot¶

Average SAT Scores per Borough Boxplot¶

Clustered Marker Map of Schools in NYC¶

Rat Sightings¶

Number of Rat Sightings per Borough Barplot¶

Average Number of Sightings per km^2 per Borough Barplot¶

Rat Sightings Heatmap in NYC¶

Rat Sightings Google HeatMap¶

###The graph doesn't show in HTML but it will run in Jupyter¶

Population Data¶

Next we want to see the population distribution in each borough and the projected population over the next 2 decades¶

Population Growth Per Borough Grouped Barplot¶

Crime Data¶

Total Number of Crimes from 1990 to 2018 for all 5 Boroughs Lineplot¶

Number of crimes per population in all 5 boroughs¶

Number of Crimes per area in all 5 boroughs¶

Proportion of Types of Crimes in 2018¶